Experimental Study on Sandwich Composites with Carbon Fiber Fabric Skin and 3D-Printed Carbon Fiber-Reinforced Polyamide Core

Abstract

Material extrusion is a widely employed additive manufacturing technique with the functional capability of fabricating solid objects or cellular structures by depositing molten thermoplastic material in successive layers according to the designed path of deposition beads. Carbon fiber sandwich composites are advanced structures, ideal in applications that require high strength and stiffness, and low weight. In the present work, sandwich composites consisting of a carbon fiber-reinforced polyamide core—3D printed with a cubic pattern at 50% infill density—and carbon fiber fabric (CFF) skin, were fabricated using the hand lay-up method and experimentally investigated. The results showed that the tensile, flexural and impact strength of the sandwich composites increased by 64.5%, 24.5% and 69.0%, respectively, compared to unreinforced 3D printed specimens with 50% infill density, and by 24.3%, 18.8% and 56.3%, respectively, compared to unreinforced 3D printed specimens with 100% infill density. In addition, a reduction in the water absorption and the density of the sandwich composites was observed. Similar results were obtained for sandwich composites with one additional internal CFF layer. This work demonstrates that this specific combination of materials and manufacturing processes can be successfully employed for lightweight, water-resistant carbon fiber sandwich structures with improved mechanical strength.

1. Introduction

Carbon fiber sandwich structures have been extensively used in automotive, aerospace, construction, energy and shipbuilding industries due to their low specific weight, high strength-to-weight ratio, enhanced bending stiffness, good energy absorption and corrosion resistance [1,2,3]. Their basic form consists of a thick, low-density core and a thin, high-strength carbon fiber-reinforced polymer (CFRP) skin firmly bonded to the outer surfaces of the core. The core can be manufactured in various forms, including porous structures made of foam or balsa wood, corrugated structures with a wave-like pattern, auxetic cellular structures, and lattice truss structures [4,5,6]. The advanced properties of carbon fiber sandwich composites are mainly attributed to the high mechanical strength of the CFRP skin, as well as the high shear and compressive strength of the core and its ability to support the skin and prevent its buckling [4]. The properties of CFRP are highly dependent on fiber orientation, fiber volume fraction, and fiber–matrix bonding. CFRPs exhibit high specific strength and stiffness, low thermal expansion, excellent fatigue resistance, and corrosion resistance.

In recent decades, a lot of research has been carried out on the design and manufacturing of cellular cores for sandwich composites. As these cellular cores exhibit complex topology and geometry, they are difficult or impractical to fabricate by conventional subtractive manufacturing. In contrast, the emerging technology of additive manufacturing (AM) allows the fabrication of highly complex structures with high accuracy and efficiency [7]. Fused deposition modelling (FDM) is a widely employed AM material extrusion process that creates objects by depositing fused thermoplastic material in subsequent layers [8]. The feedstock material, in the form of a filament, is extruded through a nozzle in a semi-molten stage and deposited in beads (also referred to as rasters) on the building platform and on subsequent layers. One of the main advantages of FDM 3D printing is the operational ease of manufacturing customized cellular core structures of two-dimensional patterns or three-dimensional unit cell geometries. Two-dimensional cellular cores are structures with a repetitive pattern in each core layer (also referred to as in-plane porosity) and are usually designed using computer-aided design (CAD) software. Some common two-dimensional core structures include honeycomb, hexagonal, re-entrant, triangular and prismatic [7,9]. Three-dimensional lattice cores are arrangements of repeated three-dimensional unit cells, such as octet, octahedral, pyramidal, X-type, Y-type, gyroid, diamond, and Kagome [7,9]. These porous structures can either be strut-based or surface-based, and their design is usually performed using mathematical algorithms and specialized software to create triple periodic minimum surfaces (TPMS) topologies [10].

A less complicated procedure to create 3D-printed cellular cores, which eliminates the extensive design and modelling process, is the determination of the infill pattern of the core structure directly in the 3D printing slicing software. Through this procedure, infill patterns can be generated as two-dimensional (e.g., lines or rectilinear, triangles or triangular, zig zag, concentric, tri-hexagonal), or as a network of three-dimensional unit cells (e.g., cubic, gyroid, cubic subdivision, cross 3D, octet, 3D honeycomb) [11,12,13].

The infill pattern is a 3D printing process parameter that plays a crucial role in determining the stiffness, strength and weight of the 3D-printed core. The maximum strength of a 3D-printed object is generally achieved when the beads are aligned with the direction of the applied loads (e.g., compression, tension, or shear) [14]. In the case of applications with loads that cause multidirectional stress, the infill pattern must ensure uniform distribution of strength in all directions. To achieve this, either three-dimensional infill patterns or two-dimensional patterns that change the direction of deposition in subsequent layers would be more suitable [15]. In addition to the infill pattern, the infill density, which determines the percentage of filling material in the inner volume, is also an important factor that affects the strength and stiffness of the 3D-printed part [16,17], as well as the printing time, the amount of material used, and the cost of the filament.

Table A1 in Appendix A presents a summary of recent experimental findings on the effect of the infill pattern on the mechanical properties of carbon fiber-reinforced polymer specimens produced by 3D printing. The infill pattern that leads to better mechanical properties is highlighted in bold. It should be noted that for a given material and fixed printing parameters, a particular infill pattern may provide higher tensile strength compared to other infill configurations; however, when the 3D-printed part is subjected to a different load, i.e., bending, this particular pattern can result in lower strength, as reported in the work of Pop et al. [18]. In addition, the behavior of the interaction between the infill pattern and other printing parameters, such as infill density, layer height, and printing speed, can be quite complex, resulting in different properties. Recent studies have shown that infill density plays a crucial role in the way the infill pattern affects the mechanical properties. In the work of Andreozzi et al. [19], the higher compression strength for PA6 CF-GF samples at 100% infill density is provided by the concentric pattern, while for samples at 50% infill density, it is provided by the grid pattern. Similar observations have been reported in the work of Ansari and Kamil [20], where the highest hardness value for PLA-CF samples at 100% infill density is provided by the grid and triangular pattern, while for samples at 75% infill density, it is provided by the tri-hexagonal pattern.

Manufacturing techniques for sandwich-type composites vary depending on application requirements and cost considerations, ranging from simpler methods such as hand lay-up and vacuum bagging to more advanced processes including vacuum infusion and compression molding. In recent years, advances in additive manufacturing (AM) technologies have further expanded the available fabrication methods. The wide availability of FDM materials and the flexibility of AM for cellular core design enable the creation of a vast range of complex sandwich-type composites. The core and skin can either be fabricated separately via 3D printing and subsequently bonded using adhesives or simultaneously fabricated in a single AM process using advanced dual-nozzle 3D printers [21,22,23].

These advances have also stimulated the development of hybrid sandwich composites that integrate 3D-printed cores with fiber-reinforced skins manufactured through conventional techniques. Consequently, an increasing number of studies have examined such sandwich composites, typically fabricated using hand lay-up or vacuum bagging. Table A2 in Appendix A summarizes recent research from the past five years focusing on these sandwich structures. Polylactic acid (PLA) is currently the most commonly employed core material due to its biodegradability, cost-effectiveness, and ease of processing via 3D printing. The mechanical properties most frequently evaluated for these sandwich composites include three-point bending strength, compressive strength, and low-velocity impact resistance, whereas tensile and impact performance have received comparatively limited attention.

In the present work, sandwich composites that combine a high-temperature carbon fiber-reinforced polyamide (PAHT-CF) core and carbon fiber fabric (CFF) skin have been fabricated. The core of the specimens was additively manufactured using an FDM 3D printer with a cubic infill pattern at 50% infill density. Using the hand lay-up method, the carbon fiber fabric impregnated with epoxy resin was bonded to either side of the core to form the skin reinforcement. An additional sandwich-type composite was fabricated with an internal CFF layer embedded at the mid-plane of the specimens. The strength performance of the sandwich composites was experimentally determined through tensile test, three-point bending test, impact test, and hardness measurements, and the results were compared to unreinforced PAHT-CF specimens 3D printed at 50% and 100% infill density. In addition, density and water absorption measurements were performed for both the unreinforced specimens and the sandwich composites.

This work contributes to the growing body of literature referring to sandwich-type structures with 3D-printed core as follows:

a.

The proposed sandwich structure, consisting of a carbon fiber fabric/epoxy resin skin and a 3D-printed PAHT-CF core, prepared using the hand lay-up technique, has not yet been experimentally benchmarked in the literature, to the best of the authors’ knowledge.

b.

A series of different tests has been conducted, including tensile, three-point bending, impact, hardness, density, and water absorption, providing a comprehensive view of the most important properties, often required in a variety of engineering applications.

c.

Although the effect of the cubic infill pattern on the properties of FDM 3D-printed samples has been investigated in the literature, it has not been studied for PAHT-CF material, to the authors’ knowledge. In this work, a preliminary investigation was conducted to evaluate three infill patterns in terms of mechanical strength and dimensional accuracy.

d.

The outcomes of this work highlight the suitability of the proposed sandwich composites as lightweight, high-strength, and water-resistant structures for relevant applications.

The rest of the paper is organized as follows. Section 2 describes the materials, fabrication process, and testing methods. Section 3 presents the results and discusses the findings, including the preliminary evaluation of different 3D-printed infill patterns and the experimental investigation of the performance of the fabricated specimens. Finally, Section 4 presents the conclusions.

2. Materials and Methods

2.1. Materials

For the additive manufacturing of the unreinforced specimens as well as the core of the sandwich specimens, the 3D printing filament Ultrafuse^® PAHT CF15 (BASF, Emmen, Netherlands) was utilized. According to the supplier’s technical data sheet [24], PAHT CF15 filament consists of a PA6-based copolymer matrix with 15% chopped carbon fibers, offering good dimensional stability, good printability, low water absorption, high temperature resistance up to 150 °C, and advanced mechanical properties. Specifically, the 3D-printed test specimens exhibit a tensile strength of 103.2 MPa (ISO 527 [25]), a flexural strength of 160.7 MPa (ISO 178 [26]), and an impact strength Izod (unnotched) (ISO 180 [27]) of 1.64 J/cm². For convenience, Ultrafuse^® PAHT CF15 will be referred to as PAHT-CF in the rest of the paper.

For the fabrication of the sandwich specimens, the EC3X-208T fabric (FTC, Douliu, Taiwan) was used as skin reinforcement. The EC3X-208T fabric is a 2 × 2 twill woven carbon fiber fabric with 3K tows oriented at 0° (warp) and 90° (weft), a nominal thickness of 0.25 mm, and a density of 208 g/m². In this type of weave, warp yarns run along the length of the fabric roll (i.e., machine direction) and weft yarns run across the width (i.e., cross direction) [28].

CFF was impregnated with epoxy resin (PRIME^TM 37, d = 1.13 g/cm³) mixed with a slow hardener (Ampreg^TM 3X) at a ratio of 100:29 by weight [29] and then was attached to the outer surfaces of the core to form the skin of the sandwich structure. Both epoxy resin and hardener were supplied by Gurit (UK) Ltd. (Newport, United Kingdom).

2.2. Fabrication Process

The CAD models of the test specimens were designed using the Autodesk Fusion software, version 2.0.20948. Their dimensions were determined considering the standards for tensile test (ASTM D638 [30]), three-point bending test (ASTM D790 [31]), and impact test Izod (ISO 180 [27]), as illustrated in Figure 1.

Figure 1. Design dimensions of specimens (in mm) for (a) tensile test, (b) impact test, and (c) flexural test.

For each mechanical test, four categories of specimens were investigated, as shown in Figure 2. The first two categories refer to the unreinforced 3D-printed specimens, while the third and fourth category refers to sandwich specimens with 3D-printed core and CFF reinforcement. A code has been assigned to each category for convenience.

Figure 2. Illustration of the fabrication process.

The unreinforced specimens (P100 and P50), as well as the core of the sandwich specimens, were fabricated using an Ultimaker S5 3D printer (Ultimaker, Utrecht, Netherlands). The Ultimaker Cura 5.7.1 software was used to set the printing parameters, slice the CAD models, simulate the FDM printing process, and generate the G code. Table 1 lists the process parameters, determined according to the supplier’s printing recommendations [24], empirical printing tests by the authors [32,33], and printing parameters reported in the literature for the same material [18,34,35,36,37].

Table 1. Printing process parameters for PAHT-CF.

Printing Parameters	Value
Layer height (mm)	0.15
Initial layer height (mm)	0.27
Raster width (mm)	0.58
Number of walls	3
Wall thickness (mm)	1.74
Number of top layers	3
Number of bottom layers	3
Top/bottom thickness (mm)	0.9
Build orientation	XY
Infill density (%)	50%, 100%
Infill pattern	Lines, Cubic, Concentric
Top/bottom layers pattern	Lines, Concentric
Infill overlap	10%
Infill flow, top/bottom flow	100%
Print speed (mm/s)	30
Initial layer speed (mm/s)	15
Extrusion head type	CC 0.6 mm
Nozzle printing temperature (°C)	270
Build plate temperature (°C)	100
Raft material	None

The infill density represents the percentage of filling material within the interior of the 3D-printed part, which is enclosed by the bottom and top layers as well as the walls. At 100% infill density, the specimens are fully solid and capable of withstanding significant loads. In our study, we focused on the 100% and 50% infill densities, as these values have been adopted in previous studies employing the same PAHT-CF material [18,34,35,36,37], enabling direct comparison with the current results. Furthermore, the 50% infill density was selected as an appropriate balance between maintaining mechanical strength and minimizing material usage.

The values for the other parameters were chosen appropriately to prevent warping and provide good adhesion and high printing accuracy. It is also worth noticing that the infill is fully enclosed within a shell consisting of a wall with a thickness of 1.74 mm, as well as three top and three bottom layers with an overall thickness of 0.45 mm on either side. As suggested by the findings of [38], the top and bottom layers enhance adhesion between the 3D-printed core and the CFRP skin and can reduce the risk of debonding during mechanical testing. Moreover, specimens at 100% infill density were fabricated with different pattern combinations for the infill and top/bottom layers. These specimens were then evaluated for mechanical strength, dimensional accuracy, and surface finish, as presented in Section 3.1.

The third category (P50R) refers to sandwich specimens that consist of the 3D-printed PAHT-CF core at 50% density and the CFF skin. The core was 3D printed with a reduced thickness of 0.5 mm, considering the thickness of the CFF skin (i.e., 0.25 mm per layer). The hand lay-up technique was adopted, and one layer of CFF impregnated with epoxy resin was attached to both the top and bottom surfaces of the 3D-printed core. In all cases, the 3D-printed core samples were placed with their principal axis (length direction) oriented along the machine direction of the fabric. Epoxy resin was mixed with the appropriate amount of slow hardener to control the exothermic reaction and prevent the epoxy from hardening too quickly.

The fourth category (P50M) refers to sandwich specimens that consist of the 3D-printed PAHT-CF core at 50% density, the CFF skin, and one internal CFF layer embedded at the mid-plane of the specimens. The core was 3D printed in two halves with a thickness of 4.62 mm each for the flexural and impact test specimens, considering the thickness of the CFF skin and the internal CFF layer (i.e., 0.25 mm per layer). Using the hand lay-up technique, a single CFF layer was impregnated with epoxy resin, placed among the halves, and joined them into one piece. Then, one layer of CFF impregnated with epoxy resin was attached to each outer surface to form the CFF skin.

The hand lay-up technique for all sandwich specimens was performed at a constant room temperature of 20 °C and a relative humidity of 50%. Prior to joining, the surfaces of the 3D-printed core were ground with sandpaper #220 to create roughness on the surface and promote adhesion. After fabrication, all sandwich specimens were fixed using clamps, left for seven days under the same environmental conditions to achieve slow curing, and then machined with simple tools to obtain their final shape. Images from the FDM 3D printing process and the hand lay-up technique are provided in Figures S1 and S2 in the Supplementary Material.

2.3. Testing Methods

A minimum of three specimens were fabricated and tested for each category (P100, P50, P50R, and P50M). The mean and standard deviation of the measured values for density, water absorption, hardness, and mechanical strength have been reported to reflect the variability in measurements and the consistency of the results.

All mechanical tests were conducted at a constant room temperature of 20 °C and relative humidity of 50%. Density and water absorption measurements were performed at a temperature of 19 °C.

Tensile and flexural strength were determined according to ASTM D638 [30] and ASTM D790 [31] standards, respectively, using a universal machine with a load cell capacity of 10 kN (±0.01 kN accuracy) and a constant crosshead displacement of 2 mm/min. During the test, the load applied to the specimen gradually varied until the failure point of the specimen, and the elongation was recorded by a Force-length meter. For the tensile test, a clamping collar was installed on the upper and lower sides to hold the specimen, while for the flexural test, a three-point bending fixture was used instead.

The impact strength of the specimens was determined according to ISO 180 [27] using the Izod method. An Avery–Denison impact testing machine was used, equipped with a 0.975 kg hammer and a built-in dynamometer that records the energy of fracture. The impact speed of the hammer was set to 3.46 m/s.

Hardness (Shore-D grade) measurements were carried out according to the standard ISO 868 [39]. Three surfaces were examined for each specimen: the top surface, the surface on the vertical side, and the bottom surface. Hardness measurements were taken at five different points on each examined surface.

The density of the specimens was determined according to ASTM D3800 [40] using the buoyancy method. A single piece of each specimen weighing 3–4 g was initially weighed in air and subsequently fully immersed in absolute ethanol (density 0.79 g/cm³) using a suspension wire. The mass of the samples was measured using an electronic analytical balance of ±0.1 mg accuracy. The density of each sample was calculated in g/cm³ using the formula:

d_i = ((m₃ − m₁)∙d_l)/((m₃ − m₁) − (m₄ − m₂)),

(1)

where d_l is the density of ethanol, m₁ is the mass of the wire in air, m₂ is the mass of the wire immersed, m₃ is the mass of the sample in air, and m₄ is the mass of the sample immersed.

The water absorption capability of the specimens was determined according to ASTM D570 [41]. Single pieces of each specimen were immersed in glass containers with distilled water after being placed in an oven at 100 °C for 1 h and then dried at room temperature. At specific intervals, the pieces were wiped to remove surface water and weighed using an electronic analytical balance of ±0.1 mg accuracy. The measurements were repeated until no significant change in mass was observed. Water absorption per unit volume was calculated in mg/cm³ using the formula:

Water absorbed = ((m_i − m₀)/V)·1000,

(2)

where m₀ is the initial mass of the sample (dry) at time t₀, m_i is the mass of the sample (immersed) at time t_i, and V is the volume of the sample, given by V = m₀/d, where d is the density of the sample.

3. Results and Discussion

3.1. Preliminary Evaluation of Infill Patterns for 3D-Printed PAHT-CF Specimens

For a given material, the maximum value of a mechanical property corresponds to a specific combination of printing parameters, and, in some cases, is significantly influenced by the interaction between the infill pattern and other process parameters, as discussed in Section 1 and summarized in Table A1 (Appendix A).

In the literature, only a limited number of studies [18,27,34,35,36,37] have investigated the properties of 3D-printed specimens fabricated with the same PAHT-CF material used in the present work. The reported results vary depending on the specific printing equipment, user-defined FDM process parameters, and the selected infill pattern (e.g., triangles, lines, concentric, zig zag, grid), making it challenging to draw direct conclusions regarding which infill configuration yields superior mechanical performance.

In this context, it was deemed necessary to conduct a preliminary study and investigate the influence of infill pattern on the mechanical strength of 3D-printed PAHT-CF specimens under the experimental processing conditions adopted in this work. Specimens were 3D printed at 100% infill density to provide maximum strength values, using different pattern combinations including three patterns for the infill (i.e., lines, concentric and cubic) and two patterns for the top/bottom layers (i.e., lines and concentric). Table 2 illustrates the four pattern combinations for the tensile specimen and the corresponding code symbols. The dimensional accuracy, surface finish, tensile strength, elongation at break, and impact strength of the 3D-printed specimens were then evaluated.

Table 2. Different combinations of patterns for the infill and top/bottom layers.

Infill Pattern		Top/Bottom Pattern	Code
Lines		Lines	P100LL
Concentric		Concentric	P100OO
Cubic		Lines	P100CL
		Concentric	P100CO

Regarding the infill, the three specific patterns—lines, concentric, and cubic—were selected due to their distinct characteristics and advantageous applications across various models and requirements. Specifically, the lines pattern is one of the simplest and fastest to print, uses minimal material, and reduces filament costs [15]. The concentric pattern takes more time to print; however, it has been reported to exhibit a strong bonding between rasters at 100% infill density [14] and high surface quality [42].

The cubic pattern creates a network of three-dimensional tilted cubes in the infill domain and provides uniform strength in all directions, which is important in engineering applications [15]. This three-dimensional pattern has recently gained popularity, and its effect on the properties of FDM samples has been evaluated in comparison with other infill patterns [13,16,43,44,45,46]. However, the cubic pattern has not, to the authors’ knowledge, been investigated to date for 3D-printed PAHT-CF specimens, unlike other patterns, such as grid [18], concentric and zig zag [34], triangles [18,32,37], and lines [18,35,36].

Regarding the top/bottom layers, the lines and concentric patterns were primarily selected to ensure continuity with the corresponding infill patterns, that is, lines and concentric. For all specimens, a bed temperature of 100 °C and a low initial layer speed of 15 mm/s were set to ensure good adhesion of the initial layer to the building platform, minimum edge warpage, and sufficient bonding among the printing beads of the bottom layers.

3.1.1. Mechanical Properties

Figure 3 shows the tensile stress–strain curves obtained for the 3D-printed PAHT-CF specimens at 100% infill density. Two specimens were 3D printed for each pattern combination and tested until failure. For all specimens, the fracture occurred along the length of the gauge.

Figure 3. Stress–strain curves and PAHT-CF specimens after tensile test.

The average values of the tensile properties are depicted in Table 3. The highest ultimate tensile strength was obtained for the Concentric/Concentric specimens (P100OO), i.e., 127.8 MPa, followed by the Cubic/Concentric specimens (P100CO), i.e., 117.3 MPa. Regarding elongation at break (% ε_max), the higher value was calculated for the P100OO specimens, followed by P100CL. However, the standard deviation values in %ε indicate that P100CL is more sensitive regarding the printing process, while P100CO is the most consistent for FDM 3D printing in terms of repeatability.

Table 3. Tensile properties and impact strength for PAHT-CF specimens.

	Tensile Properties		Impact Strength (J/cm2)
Specimens	σmax (ΜPa)	% εmax	Face-Up	Side
P100LL	93.4 ± 0.5	5.7 ± 0.3	-	-
P100OO	127.8 ± 3.4	7.0 ± 0.7	3.29 ± 0.08	2.63 ± 0.11
P100CL	96.5 ± 13.0	6.3 ± 1.0	2.65 ± 0.47	2.56 ± 0.09
P100CO	117.3 ± 4.5	4.8 ± 0.2	1.99 ± 0.16	2.51 ± 0.14

The higher tensile strength of the P100OO specimens can be attributed to the concentric infill pattern and the bead deposition paths. In this pattern, the tensile load is applied along the larger dimension of the 3D-printed specimen (the X-axis), which also aligns with the direction of the deposited beads. Additionally, the higher elongation of P100OO specimens can be explained since strain depends on the orientation of the beads (raster angle), with greater strain occurring when the raster direction aligns with the X-axis, as reported in [47]. Unlike the concentric pattern, in the lines and cubic patterns, the printing beads are deposited at 45°/135° raster angles. In these cases, the tensile load is not aligned with the deposited beads, resulting in lower strength. As observed in Table 3, the tensile strength for the PAHT-CF specimens at 100% infill density was found to be close to the value specified by the supplier (i.e., 103.2 MPa [24]), as well as to values reported in similar experimental studies [34,36]. These small variations are likely due to the specific printing parameters used in this experimental work, which influence the mechanical strength of the specimens.

Impact strength was calculated as the absorbed energy per unit cross-sectional area of the specimens when the impact force was applied vertically to the top surface (face-up) and vertically to the side surface (side) of the 3D-printed specimens. Figure 4 shows the direction of the impact force on the surfaces of the P100CO impact specimen.

Figure 4. Direction of impact force for the test specimen with a cubic pattern. The top layer has been removed to display the infill.

Two unnotched specimens were tested for each pattern combination, and the results were obtained by averaging the data (Table 3). It can be observed that when the impact force was applied vertically to the top surface, the specimens with the concentric pattern (P100OO) exhibited higher impact strength, i.e., 3.29 J/cm², than those with the cubic pattern, i.e., 1.99 J/cm². However, when the impact force was applied vertically to the side of the specimens, no significant difference in impact resistance was observed between the concentric and cubic infill patterns.

One possible explanation for the differing impact resistance of the infill patterns with respect to the direction of the applied force is the quality of inter-bead and inter-layer bonding. Since the specimen is 3D printed along the Z axis, the adjacent beads are deposited in the XY plane while the layers are built successively along the Z axis. In the lines and concentric patterns, a 10% infill overlap was set, which created negative air gaps between the adjacent beads in each layer. This potentially improved inter-bead bonding, and when the impact force was applied perpendicularly to the top surface, it limited crack propagation through both the top layers and the infill. On the contrary, the beads in the cubic pattern are not deposited as adjacent parallel lines in the XY plane and cannot overlap. As a result, small air gaps form along each layer, leading to a slightly lower impact resistance.

When the impact force is applied perpendicularly to the side surface (i.e., vertically to the deposition of the 66 layers), it appears that the adhesion between successive layers is not particularly affected by the infill pattern, and, therefore, no change in impact resistance is observed. In this case, the bonding between successive layers may depend more on printing parameters—layer height, nozzle temperature, and print speed—than on the infill pattern.

3.1.2. Dimensional Accuracy and Surface Quality

Dimensional accuracy is a crucial factor for applications that require strict tolerances and high functionality. The achievement of the design dimensions mainly depends on the printing process parameters and the respective material behavior during the 3D process, and is affected by 3D printing issues (e.g., warping, shrinkage, inconsistent extrusion). To evaluate the precision of the different pattern combinations, dimensional measurements of the 3D-printed tensile specimens were performed using a micrometer with an accuracy of 0.01 mm. Three measurements were taken for each dimension, and the corresponding mean values and standard deviations are reported in Table 4.

Table 4. Dimensional measurements of the 3D-printed PAHT-CF tensile specimens (in mm).

Specimens	Length	Width	Thickness
P100LL	100.39 ± 0.09	10.41 ± 0.05	3.17 ± 0.01
P100OO	100.18 ± 0.06	10.23 ± 0.04	3.06 ± 0.01
P100CL	100.71 ± 0.01	10.48 ± 0.30	3.14 ± 0.09
P100CO	100.31 ± 0.01	10.22 ± 0.22	3.05 ± 0.01
CAD model	100.00	10.00	3.00

In all cases, there was a positive deviation in the size of the 3D-printed specimens in all three dimensions compared to the design dimensions of the CAD model. The Concentric/Concentric combination demonstrated the best dimensional accuracy a both the X and Y axes (specimen length and width), followed closely by the Cubic/Concentric combination. Regarding the Z-axis (thickness of the specimen), the Cubic/Concentric and the Concentric/Concentric specimens showed the highest dimensional accuracy. The Lines/Lines and Cubic/Lines pattern combinations exhibited greater dimensional deviations, likely due to the elevated temperatures developed in the adjacent beads in the top and bottom layers, which led to an expansion of the specimen dimensions.

Surface quality is also an important indicator of the appearance, quality and functionality of a 3D-printed object. It corresponds to surface imperfections such as visible grooves, grainy texture, small deviations and irregular patterns, and depends directly on the printing process parameters and the material used [48]. In all 3D-printed specimens, the line pattern resulted in poor surface quality of the top/bottom layers with noticeable imperfections such as visible grooves and low continuity between the infill and the walls. On the other hand, the concentric pattern provided a uniform and smooth surface finish on the top and bottom layers.

3.1.3. Infill Pattern Considerations

Some additional considerations arising from the FDM printing process that could potentially explain the effect of the infill pattern on the quality and mechanical strength of PAHT-CF 3D-printed parts are as follows:

In the concentric infill pattern (Figure 5a), the deposition path follows the contours of the outer walls from the outside inwards, so each bead has enough time to cool before the next adjacent bead is deposited. In addition, since this two-dimensional pattern is the same at each layer, small air gaps are created between the beads, which extend vertically in the internal volume of the printed object. These vertical passages of voids are prone to crack propagation. Setting the infill overlap to 10% can further improve the inter-bead bonding. The concentric pattern is preferred for axial loads in one direction, particularly parallel to the largest dimension (as in the case of the tensile test specimens).

Figure 5. (a) The concentric infill pattern; (b) The cubic infill pattern. Walls and top/bottom layers have been removed to display the infill.

In the lines infill pattern, the material is deposited in parallel lines that alternate their direction from layer to layer (45° for the first layer and 135° for the next layer). The 45°/135° raster angles provide a more equal distribution of strength in both the X and Y directions, compared to the 0°/90° raster angles. However, in each layer, the length of the deposition lines is very short, and beads are deposited next to each other very quickly, resulting in very little time to cool properly. The high temperature that develops between adjacent beads can cause a slight enlargement of the 3D-printed object, and this can be strongly observed in objects with a smaller size (for example, in the case of the tensile specimens of 3 mm thickness and 10 mm width).

In the cubic infill pattern (Figure 5b), a network of three-dimensional cubes with one corner facing down is formed. The deposition path of the beads is different in each layer in the XY plane, resulting in the creation of air gaps at different locations in each layer. In this way, there are no continuous vertical passages of voids along the specimen’s height where a crack could easily propagate. Moreover, the nozzle moves over longer distances along the entire XY plane in a random order, rather than in short parallel lines created in the lines pattern. These longer paths allow for proper cooling of the deposited material, as opposed to the line pattern. In addition, beads are not parallel to only one stress axis (as occurs in the concentric pattern) but are deposited in 45°/135° raster angles, providing uniform stress distribution along the X and Y directions.

Based on the 3D printing results and mechanical tests, the Concentric/Concentric pattern combination resulted in 3D-printed samples with the highest strength and good quality (dimensional accuracy and visual surface finish). However, the concentric pattern is more suitable for structures with a high length-to-width ratio, such as arms or beams, where the tensile load is applied to the longer dimension. Instead, the cubic pattern is considered more suitable to withstand stress in all three axes, which is generalized to most real-world structures, including biomedical devices, automotive components, aerospace components, and marine structures. Regarding the top/bottom layers, the concentric pattern exhibits a smoother surface finish without imperfections, compared to the line pattern. Under this reasoning, the Cubic/Concentric pattern combination was considered the most appropriate for 3D printing the unreinforced PAHT-CF specimens, as well as for the core of the sandwich specimens.

3.2. Experimental Investigation of the Performance of the Fabricated Specimens

3.2.1. Density and Water Absorption Measurements

The density was determined in accordance with ASTM D3800 [40] and the corresponding mean values and standard deviations for categories P100, P50, and P50R are presented in Figure 6.

Figure 6. Density values in g/cm³ for the unreinforced specimens (P100, P50) and the sandwich specimens with 3D-printed core and CFF skin (P50R).

The density of the unreinforced PAHT-CF specimens at 100% density (P100) was found to be lower than the density of the filament, as specified by the suppliers (i.e., 1.203 g/cm³ [24]). This variation can be attributed to the porosity created during the printing process, i.e., the voids that are generated between the deposited beads and layers. Also, as expected, a decrease in density of 18.26% (average value) was observed for the P50 unreinforced specimens compared to the P100 unreinforced specimens.

Regarding the sandwich specimens with a 3D-printed core and CFF skin (P50R), they exhibited a slight increase in density compared to the P50 specimens. This increase is mainly attributed to the higher density of CFF/epoxy resin, which replaced the corresponding volume of PAHT-CF in the P50R specimens, compared to the density of the 50% infilled PAHT-CF material. It is also interesting that the density of the P50R sandwich specimens was found to be even lower than that of the solid P100 specimens. This is a particularly important finding, showing that this specific choice of materials and fabrication technique can provide structures with reduced weight.

Water absorption was determined according to the standard ASTM D570 [41] for three different samples for each of the categories P100, P50, and P50R. The typical curves of water absorption expressed as mg H₂O per 100 mg are shown in Figure 7. In addition, the average values of water absorbed by the specimens in a 35-day immersion period in distilled water are reported in Table 5.

Figure 7. Water absorbed per unit volume for the unreinforced specimens (P100, P50) and the sandwich specimens with 3D-printed core and CFF skin (P50R).

Table 5. Average values and standard deviation for water absorbed in a 35-day period.

Specimens	Mass (g)	Water Absorbed (mg/cm3)	St. Dev.	Water Absorbed (%)	St. Dev.
P100	3.190	131.671	33.589	13.048	3.280
P50	2.626	199.638	38.862	23.852	4.643
P50R	3.110	118.973	25.930	13.383	2.917

Regarding the P100 unreinforced specimens, the water absorption percentage was found to be equal to 3.20% after a 24 h period (1 day) of immersion in distilled water. This percentage of water absorption is attributed to the presence of voids between the beads and layers created through the 3D printing process. This percentage is consistent with findings in similar experimental works [47], where the water absorption for PA6 specimens emerged in distilled water at room temperature for 24 h was calculated to be approximately 2.8–2.9%. Regarding the P50 unreinforced specimens, it was found that reducing the infill density by 50% leads to a significant increase in water absorption of 51% after a 35-day immersion period, compared to the P100 specimens, which was expected.

It was also observed that the P50R sandwich specimens exhibited a significantly lower tendency to absorb water compared to the P50 specimens, even though both categories have the same infill density of 50%. It is also interesting that the water absorption tendency of P50R sandwich specimens was found to be even slightly lower than that of the solid P100 specimens at 100% infill density. This is mainly attributed to the replacement of the corresponding volume of PAHT-CF in the specimen with the epoxy resin skin, which exhibits high water resistance.

Furthermore, the CFF skin hinders water ingress into the internal structure of the sandwich specimens. Despite the 50% reduction in PAHT-CF core density, which generates significant air gaps and increases the composite’s void volume fraction, the CFF skin acts as a protective barrier that limits water absorption by the cellular cubic core.

The reduced density and water absorption tendency of the proposed sandwich specimens with 3D-printed PAHT-CF core and CFF skin may prove beneficial for applications in the water sports and shipbuilding sectors, where lightweight and water-resistant structures are essential. The standard deviation for water absorption measurements regarding the P50R sandwich specimens was also found to be the lowest, indicating good reproducibility of the experimental results.

3.2.2. Hardness Measurements

Shore D hardness measurements were performed on the P100 unreinforced specimens and P50R sandwich specimens according to standard ISO 868 [39]. The corresponding values, standard deviation, and number of points (n) are presented in Table 6.

Table 6. Shore D Hardness for the unreinforced specimens (P100) and the sandwich specimens with 3D-printed core and CFF skin (P50R).

Specimens	Surface	Hardness	St. Dev.	n
P100	face-up	79.1	1.8	10
P100	face-down	79.5	3.0	10
P100	side	71.0	2.6	10
P50R	face-up	83.7	4.1	10
P50R	face-down	82.3	3.6	10

Regarding the unreinforced PAHT-CF specimens, Shore D hardness was measured on three surfaces: (i) the top surface (face-up), (ii) the bottom surface that was in contact with the glass build plate (face-down), and (iii) the surface on the vertical side of the specimen (side).

According to the results for the unreinforced specimens, the hardness measured on the top and bottom surfaces was higher than that measured on the side surface. This indicates a good bonding between the beads in the top/bottom layers, which can be mainly attributed to the concentric pattern and the user-specified overlap parameter. In addition, the mean hardness was found to be higher than the hardness of the PAHT-CF, as specified by the supplier (i.e., 72 [24]). This variation can also be attributed to the user-defined printing parameters, which enhanced the hardness on the top and bottom surfaces of the 3D-printed specimens. Furthermore, the hardness on the side surface, i.e., perpendicular to the deposition of the 66 layers, was found to be lower, but close to the supplier’s value, indicating also good inter-layer bonding.

For the P50R sandwich specimens, hardness was measured on two surfaces: (i) the top surface of the CFF skin (face-up) and (ii) the bottom surface (face-down) that was in contact with the glass build plate (face-down). It was found that the CFF skin exhibited higher hardness compared to the PAHT-CF material, corresponding to an increase of 4.6% (on average).

3.2.3. Tensile Strength

Figure 8 depicts the tensile stress versus strain curves of the P100 and P50 unreinforced specimens as well as the P50R sandwich specimens. The results showed that the P50R sandwich specimens exhibited the highest ultimate tensile strength of 145.8 MPa, which corresponds to a remarkable increase of 64.5% compared to the 50% density PA-CF specimens, i.e., 88.6 MPa, and an increase of 24.3% compared to the 100% density PA-CF specimens, i.e., 117.3 MPa. In addition, the P50R sandwich specimens also exhibit the highest elongation at break, indicating a more elastoplastic behavior compared to unreinforced specimens. This typical behavior is primarily attributed to the high tensile strength of the CFF skin, which enables it to carry almost all the in-plane tensile load, allowing the structure to withstand significantly higher loads without failure.

Figure 8. Tensile stress—strain curves of the fabricated specimens.

Figure 9 shows the five P50R sandwich specimens after fracture. Except for specimen no. 4, in all other specimens, an initial brittle fracture occurred in the CFF skin along the gauge length, while the fractured surface of the PAHT-CF core appeared rough, with sharp peaks located in the porous regions, indicating ductile deformation. This tensile behavior aligns with typical fracture patterns reported in the literature for 3D-printed PAHT-CF samples [34,36], as well as for sandwich structures with polymer matrix cores and fiber-reinforced polymer skins [49].

Figure 9. Sandwich specimens (P50R) after fracture. The red line in the specimen no. 4 indicates the fracture line.

The corresponding stress–strain curves for the acceptable specimens (except specimen no. 4) are depicted in Figure 10.

Figure 10. Tensile stress—strain curves of the P50R sandwich specimens.

3.2.4. Impact Strength

To determine the impact strength of the P100, P50, and P50R specimens, the impact force was applied in two different ways: (i) perpendicular to the top surface of the specimens (face-up), and (ii) perpendicular to the side surface of the specimens (side). In addition, the R50M sandwich specimens, which consist of the 3D-printed core at 50% density, the CFF skin and one additional internal CFF layer, were also tested with the impact force perpendicular to the top surface (face-up). The corresponding mean values and standard deviations are depicted in Figure 11.

Figure 11. Impact strength of the fabricated specimens with respect to the impact force direction.

The results indicate that CFF skin reinforcement leads to a significant increase in the impact strength of the sandwich specimens, as illustrated in Figure 11. In particular, when the impact force was applied vertically to the side surface, the CFF skin reinforcement enhanced the impact strength of the P50R sandwich specimens by 21.3% compared to the P50 unreinforced specimens, while no difference was observed with the P100 specimens. On the other hand, when the impact force was applied vertically to the top surface (face-up), the impact strength of the P50R specimens increased by 69.0% compared to the P50 unreinforced specimens, and by 56.3% compared to the P100 unreinforced specimens. A similar impact strength was determined for the P50M sandwich specimens.

It can be clearly observed that although the density of the core has decreased to 50%, the sandwich structure showed improved impact resistance, even higher than that of the P100 specimens. This is mainly attributed to the carbon fiber skin, which distributes the impact load over a larger area and reduces local stress concentration in the core, allowing the absorption of larger amounts of energy, thus increasing the overall impact strength.

Figure 12 shows the crack propagation generated during the impact test of the fabricated specimens. Regarding the P100 and P50 unreinforced specimens, fracture occurred abruptly with minimal plastic deformation, producing rough surfaces with sharp peaks and visible voids (Figure 12a,b). This fracture behavior agrees with previous studies [34,35,36], where SEM images of fractured 3D-printed PAHT-CF specimens (at 100% infill density) revealed that the carbon fibers were highly aligned along the deposition paths, and fracture was primarily governed by fiber breakage and pull-out, with the average diameter of the pulled-out fibers ranging from approximately 7 to 8 μm.

Figure 12. Crack propagation of impact test specimens: (a,b) P50 specimen after face-up impact test and cross-section of the fractured surface; top view (c) and front view (d) of a P50R specimen completely fractured after side impact test; top view (e) and front view (f) of a P50R specimen with skin detachment after face-up impact test; (g,h) P50M specimens after face-up impact test.

Regarding the P50R sandwich specimens, it was observed that the direction of the impact force has a significant effect on the failure mode. Figure 12c,d depicts a P50R specimen subjected to a side impact force. Both the core and skin of the specimen fractured suddenly along the same crack path, resulting in complete failure. For the P50R specimens subjected to a face-up impact force (i.e., perpendicular to the CFF skin), most of them exhibited brittle fracture initiating in the skin and propagating through the core. However, in a few specimens, as shown in Figure 12e,f, the core and the upper CFF skin experienced complete fracture, while the lower CFF skin detached from the core.

Regarding the P50M sandwich specimens, the internal CFF layer provided additional strength to the sandwich structure. Figure 12g shows a P50M specimen subjected to a face-up impact force, in which the two halves of the core and the internal CFF layer cracked. However, the specimen did not separate but was held together by the upper CFF skin. Figure 12h shows another P50M specimen, in which both the lower CFF skin and the internal CFF layer have detached from the core in the region of the specimen located outside the testing machine’s holder. Delamination occurred between the internal CFF layer and the two halves of the core, while that specimen remained held together by the upper CFF skin reinforcement.

3.2.5. Flexural Strength

To determine the flexural strength of the P100, P50, and P50R specimens, the bending load was applied in two different ways: (i) perpendicular to the top surface of the specimens (face-up), and (ii) perpendicular to the side surface of the specimens (side). The R50M sandwich specimens were also tested with the bending load perpendicular to the top surface (face-up). The corresponding mean values and standard deviations are presented in Figure 13.

Figure 13. Flexural strength of the fabricated specimens with respect to the bending force direction.

As can be observed, the CFF skin reinforcement enhances the flexural behavior of the sandwich specimens and leads to higher flexural strength compared to that of the unreinforced specimens. In particular, when the bending load was applied vertically to the top surface (face-up), the flexural strength of the P50R sandwich specimens increased by 24.5% compared to the P50 unreinforced specimens, and by 18.8% compared to the P100 unreinforced specimens. A similar flexural strength was determined for the P50M sandwich specimens. The enhancement of the flexural strength of the sandwich specimens is mainly attributed to the high modulus of elasticity of the carbon fiber skin, which carries most of the bending loads (tension/compression) and allows the core to resist shear loads. On the other hand, when the bending force was applied vertically to the side surface, the CFF skin reinforcement did not provide any significant increase in flexural strength compared to the unreinforced specimens.

Figure 14 shows the crack propagation generated during the three-point bending test of the specimens.

Figure 14. Crack propagation of flexural test specimens: (a) P50 specimen after face-up bending test; (b) P50R specimens after side bending test; (c,d) P50R specimens after face-up bending test; (e,f) P50M specimens after face-up bending test.

The fracture of the P100 and P50 unreinforced specimens occurred abruptly, forming a rough fracture surface with sharp peaks, consistent with the typical brittle fracture behavior of PAHT-CF material (Figure 14a).

Regarding the P50R sandwich specimens, the direction of the bending load significantly affected the failure mode of the core and the CFF skin. In the case of the side-bending load (Figure 14b), both the core and the CFF skin fractured along the same crack propagation path. In the case of face-up bending load (i.e., perpendicular to the CFF skin), some specimens exhibited cracking in both the core and the lower CFF skin without complete separation (Figure 14c), while in others, the core and upper CFF skin fractured completely, whereas the lower CFF skin detached from the core (Figure 14d).

Regarding the P50M sandwich specimens, the CFF skin and, in particular, the internal CFF layer provided additional strength and rigidity to the sandwich structure and restricted fracture. In most P50M specimens, the lower CFF skin and the lower half of the core fractured; however, the specimen did not fully separate, as it was held together by the internal CFF layer and the upper CFF skin (Figure 14e). In a limited number of specimens, the lower CFF skin detached from the core (Figure 14f), probably due to insufficient skin-core interface adhesion. Delamination was also observed between the layers of the 3D-printed core as well as in the skin/core interfaces.

3.2.6. Consideration of Quality-Related Issues

The hand lay-up technique is a cost-effective, widely used conventional manufacturing method for sandwich composites. However, it depends heavily on the user’s experience, and quality issues may arise, including skin debonding, undesirable dimensional and geometrical deviations, as well as the formation of wrinkles, bubbles, cracks, or other defects.

Figure 15a illustrates the side surface of a P50R specimen. It can be clearly observed that the resin-impregnated carbon fabric exhibits poor adhesion to the lower surface of the 3D-printed core. Figure 15b depicts the side surface of a P50M specimen, where the CFF skin and the internal CFF layer are non-uniform and do not maintain consistent thickness along the length of the specimen. This variation in thickness can be attributed to factors such as the amount of resin penetrating the fabric, the uneven clamping of the specimens during the epoxy curing process, or the deformation occurring after unclamping.

Figure 15. (a) Side surface of P50R specimen (layer delamination during 3D printing is marked with red); (b) Side surface of P50M specimen with CFF skin and internal CFF layer.

Skin detachment or debonding often occurs in bending tests of sandwich composites, as has been reported in the literature [4,50,51]. Ridlwan et al. [50] fabricated sandwich composites consisting of (a) 3D-printed PC core and CFRP skin, and (b) 3D-printed PLA core and GFRP skin, using the hand lay-up method. During the bending test, the interface strength between the 3D-printed core and the skin could not withstand the stress that occurred, resulting in a debonding failure. According to the authors, when the crack reached the skin, the bending stress was much greater than the shear stress, so the crack changed in the horizontal direction, and debonding failure occurred. Alshaer and Harland [51] fabricated sandwich beams with CFRP face sheets and 3D-printed PA12 core structures, which underwent face/core debonding during three-point bending tests. Although in-plane shear stresses were considered minimal at the core/face interface, skin separation was attributed to the tensile/compressive stresses generated at the small areas of poor bonding or small cracks. Also, according to Zhao et al. [4], the higher compressive stress theoretically develops in the middle of the skin, and this location is most prone to debonding. However, since the uniformity at the interface between the skin and the core is not ensured, the location at which debonding occurs is random.

Apart from mechanical stresses at the core/skin interface, a fundamental cause of skin detachment is the incompatibility between the two polymer matrices, i.e., the thermoplastic PAHT-CF core and the thermoset epoxy skin. These dissimilar materials do not form strong interfacial bonds, resulting in poor adhesion and delamination.

Possible solutions to reduce CFF skin detachment include applying epoxy adhesive to the core/skin interface and optimizing curing parameters. However, due to the inherent thermoplastic/thermoset mismatch, additional measures may be necessary, such as surface treatments (e.g., plasma or chemical etching) or the use of coupling agents to enhance compatibility between the thermoset epoxy and the thermoplastic core.

3.2.7. Influence of Process Parameters on Mechanical Performance

The mechanical strength of the 3D-printed PAHT-CF specimens depends largely on the characteristics of the filament—such as crystallization behavior, melt flow rate, fiber content, and orientation—as well as on printing parameters including infill pattern and density, layer height, raster width, printing speed, and nozzle temperature [8].

Τhe PAHT-CF filament used in this study consists of a high-temperature polyamide matrix reinforced with 15% chopped carbon fibers. Carbon fibers are homogeneously distributed in the filament and highly oriented along the extrusion flow direction and deposition paths, as has been reported in similar studies [34,35,36,52].

In the present work, all specimens have been 3D printed in XY orientation; thus, carbon fibers are oriented along the principal direction (length). This fiber alignment strongly influences the mechanical response under different loading conditions. During the tensile test, the load is transferred to the carbon fibers along the same orientation, resulting in a significantly higher strength compared to that of unreinforced polyamide. In addition, in impact and flexural test specimens, the orientation of the carbon fibers is perpendicular to the direction of the load and crack growth. At higher fiber contents (e.g., 8 wt%), the perpendicular carbon fibers hinder crack propagation, contributing to the enhancement of impact strength [53].

In addition to fiber orientation, the microstructural quality of the 3D-printed samples is also reflected in their mechanical performance. During the 3D printing process, air gaps are formed between adjacent deposition beads, with their size influenced by parameters such as layer height, raster width, nozzle temperature, printing speed, and infill pattern [54]. Reducing layer height and raster width has been shown to minimize voids, as the bead cross-section becomes more rectangular with smaller rounded corners [14]. In some cases, a negative air gap is preferred to promote overlapping, thereby improving adhesion and overall structural integrity [54,55].

Furthermore, the bonding quality and void formation depend on the nozzle temperature and printing speed, which must be appropriately balanced. A higher nozzle temperature improves material fluidity, fusion between adjacent beads, and inter-raster bonding strength [8,14]. As reported by Syrlybayev et al. [56], fusion occurs when the extruded material remains above the glass transition temperature. A slower printing speed helps the material maintain this temperature, enhancing bonding between layers. Conversely, very high temperatures combined with fast printing speeds hinder the cooling of deposited beads before a new layer is applied, leading to excessive bead overlap, part expansion, geometric deviations, and poor surface quality. In contrast, low printing temperatures may reduce dimensional deviations but can cause incomplete melting, positive air gaps, and weak interlayer adhesion.

Table 7 presents the mechanical strength of the P100 and P50 specimens alongside values reported in experimental studies on 3D-printed specimens made from the same PAHT-CF material. The observed differences can be attributed to variations in the printing process parameters, such as nozzle and build plate temperatures, layer thickness, infill pattern, and printing speed.

Table 7. Comparison of the mechanical strength of PAHT-CF specimens between this study and previously published works.

Material	Infill Pattern	Tensile Strength	Impact Strength	Flexural Strength	Reference
PAHT-CF at 100% infill density	lines	117.1			[36]
	zig zag (±45°)	106.6			[34]
	lines (90°/0°)	~77.4		~162	[35]
	cubic	117.3	1.99	115.8	this study
PAHT-CF at 50% infill density	lines	59.4		107.6	[18]
	triangles	59.2		108.2	[18]
	grid	57.6		104.9	[18]
	cubic	88.6	1.84	110.5	this study

Regarding the proposed sandwich composites comprising a PAHT-CF core and CFF skin, to the best of the authors’ knowledge, no experimental studies in the literature have benchmarked composites using the same material combination.

Table A2 (Appendix A) summarizes recent studies on sandwich composites consisting of 3D-printed cores and fiber-reinforced polymer skins, including a few that employ carbon fiber-reinforced polymer (CFRP) skins. Most of these studies have focused on three-point bending, compression, and low-velocity impact tests, while tensile properties or impact toughness have remained relatively underexplored.

Table 8 presents the flexural strength of the P50R sandwich specimens along with values reported in experimental studies on conventionally manufactured sandwich composites with CFRP skins and polymer-based cores. Due to differences in materials, core geometries, and process parameters, these values are not directly comparable but are provided for general reference. Specifically, Muralidharan et al. [57] examined sandwich composites with five different PLA core geometries—namely, honeycomb, X-shape, tubular hollow, triangular, and tubular solid—and CFRP skins, reporting flexural strengths ranging from 25.46 to 100.97 MPa. Forés-Garriga et al. [7] studied composites with PEI cellular cores and CFRP skins, reporting flexural strengths ranging from 3.8 to 34.3 MPa for core densities between 5% and 30%. Similarly, Alshaer and Harland [51] investigated sandwich composites with different PA12 core geometries and CFRP skins, reporting ultimate flexural strengths ranging from 2 to 10 MPa for core densities between 0.327 and 0.561 g/cm³.

Table 8. Flexural strength of P50R sandwich specimens and corresponding values from selected experimental studies.

Core Material	Skin Material	Manufacturing Process	Flexural Strength (MPa)	Reference
PA12	CFRP	3D printing/hand lay-up	2–10	[51]
PEI	CFRP	3D printing/vacuum bagging	3.8–34.3	[7]
PLA	CFRP	3D printing/hand lay-up	25.46–100.97	[57]
PAHT-CF	CFF	3D printing/hand lay-up	137.6	this study

The primary contribution of this paper, compared to previous studies, is the development of a novel sandwich-type composite comprising a 3D-printed PAHT-CF lattice core, which provides high tensile strength and stiffness, and a CFF skin (carbon fiber fabric in epoxy resin), which enhances the overall flexural strength and impact toughness of the composite. Moreover, the cubic core was fabricated by defining its structure directly in the 3D printing software, eliminating the need for an extensive design process. In addition, the use of the hand lay-up technique provided a simple and cost-effective approach for producing sandwich-type composites.

4. Conclusions

The flexibility of material extrusion AM technology to produce complex cellular core geometries, along with the wide range of usable materials, enables the fabrication of sandwich-type structures tailored to specific application requirements. This study demonstrates a sandwich-type composite that combines a PAHT-CF cubic lattice core 3D printed at 50% infill density and a carbon fiber fabric in epoxy resin skin. This material combination produces lightweight, water-resistant structures with enhanced mechanical performance, including increased strength and stiffness. The following are the most important conclusions of this study:

• The results obtained in the preliminary evaluation (Section 3.1) indicate that the infill pattern significantly affects the tensile and impact strength of 3D-printed parts. Of the three examined infill patterns, i.e., lines, concentric and cubic, the concentric pattern should be preferred for structures where tensile strength is critical (e.g., arms or beams), while the cubic pattern is more suitable for structures subjected to non-uniform loading.
• The PAHT-CF core material enhanced the overall tensile strength and stiffness of the sandwich structure. This improvement is attributed to the high strength chopped carbon fibers embedded in the polyamide matrix, oriented along the deposition beads, that increase resistance to crack propagation and reduce the likelihood of failure.
• The CFF skin (carbon fiber fabric in epoxy resin) exhibits high mechanical strength and enhances the overall flexural strength and impact toughness of the fabricated sandwich composites. The reduced water absorption of the sandwich composites is mainly due to the water resistance of the epoxy resin and depends also on the skin thickness, which hinders water ingress into the cellular cubic core.
• PAHT-CF is easy to process via 3D printing, provided that sufficient print tests have been conducted to ensure the interaction between printing parameters and final properties. Since temperature and humidity affect the results, using a closed chamber to maintain stable conditions during 3D printing is recommended.
• The proposed fabrication process, combining 3D printing of the core and hand lay-up technique, is relatively slower than other methods for sandwich structures, such as compression molding or advanced 3D printing that integrates core and skin fabrication. Therefore, it is more suitable for small-scale or customized products.
• For larger sandwich structures, the limited build volume of FDM 3D printers can be addressed by dividing the core into smaller parts. The hand lay-up technique can then be used to gradually attach the carbon fiber fabric and join all parts with epoxy adhesive.

Future research could focus on (a) investigating the effect of different 3D-printed cellular core structures on the mechanical strength of sandwich composites; (b) fabricating sandwich composites with a 3D-printed core and CFF skin using pressure vacuum bagging to reduce geometric deviations; and (c) incorporating surface treatment to the thermoplastic core or coupling agents to improve adhesion at the core–skin interface.

In a broader context, future studies should aim to investigate and predict the performance of sandwich composite structures by developing optimal combinations of materials and core configurations capable of meeting a wider range of critical, application-specific requirements (e.g., tensile strength, modulus of elasticity, flexural strength, bending stiffness, impact toughness, low-velocity impact performance, temperature and chemical resistance). Additionally, further research should analyze larger experimental datasets and examine the interactions between process parameters to better understand the mechanisms governing the structural performance of sandwich composites.